Machine Learning (ML) Engineering Team
(Redirected from ML engineering team)
Jump to navigation
Jump to search
A Machine Learning (ML) Engineering Team is a back end engineering team that delivers ML-based software systems.
- Context:
- It can (often) have an ML Engineering Technical Lead.
- It can (typically) have ML Engineers.
- It can include or work with Data Scientists (who specialize in exploratory analysis).
- It can leverage an ML Development & Deployment Platform System to streamline the process of model development, training, and deployment.
- It can follow ML Engineering Development Lifecycle.
- It can integrate machine learning models into software products or services, ensuring their compatibility with existing systems and infrastructure.
- It can collaborate with other technical teams, such as data engineers and software engineers, to build end-to-end machine learning solutions that meet the requirements and constraints of specific use cases.
- It can monitor the performance of deployed models, iterating on them when necessary.
- It can work with the ML Ops Team to maintain the overall health and reliability of the machine learning system.
- …
- Example(s):
- a team that deliver's Netflix's Movie Recommender System Team.
- a team that deliver's Spotify's Song Recommender System Team.
- a team that deliver's PlayStation's Game Play Recommender System Team.
- …
- Counter-Example(s):
- See: ML-based Project, ML Engineering Team Culture, Software Engineering Team.
References
2023
- chat
- An ML Engineering Team is a specialized software product engineering team that focuses on designing, building, and deploying Machine Learning Models to create innovative solutions or enhance existing products and services.
- Also known as: Machine Learning Engineering Team, ML Development Team.
- It can leverage the ML Platform system managed by the ML Ops Team to streamline the process of model development, training, and deployment.
- It can work closely with Data Scientists to refine models, conduct feature engineering, and optimize algorithms for real-world applications.
- It can integrate machine learning models into software products or services, ensuring their compatibility with existing systems and infrastructure.
- It can collaborate with other technical teams, such as data engineers and software engineers, to build end-to-end machine learning solutions that meet the requirements and constraints of specific use cases.
- It can monitor the performance of deployed models, iterating on them when necessary, and working with the ML Ops Team to maintain the overall health and reliability of the machine learning system.
- Associated concepts: Data Science, Data Engineering, Model Deployment, Model Monitoring, ML Platform Ops Team
2015
- (Sculley et al., 2015) ⇒ D. Sculley, Gary Holt, Daniel Golovin, Eugene Davydov, Todd Phillips, Dietmar Ebner, Vinay Chaudhary, Michael Young, Jean-Francois Crespo, and Dan Dennison. (2015). “Hidden Technical Debt in Machine Learning Systems.” In: Proceedings of the 28th International Conference on Neural Information Processing Systems (NIPS 28).
- QUOTE: ... Cultural Debt. There is sometimes a hard line between ML research and engineering, but this can be counter-productive for long-term system health. It is important to create team cultures that reward deletion of features, reduction of complexity, improvements in reproducibility, stability, and monitoring to the same degree that improvements in accuracy are valued. In our experience, this is most likely to occur within heterogeneous teams with strengths in both ML research and engineering. ...